بهینه سازی بازوی پایین یک سیستم تعلیق خودرو برای کاهش سر و صدای جاده توسط تجزیه و تحلیل حساسیت

This study analyzed characteristics of road noise using vehicle tests and identified the 200–230 Hz range as the most important frequency for road noise reduction. Moreover, vibration sources in the vehicle suspension system were identified through transfer path analysis and coherence analysis. Using a finite element model of a vehicle suspension system, sensitivity analysis was performed to determine sensitive design factors. In order to achieve noise reduction using sensitivity analysis, the lower arm of the vehicle suspension system was found to be the most important design variable. For design optimization, we employed a robust and efficient sequential approximate optimization method, named PQRSM (Progressive Quadratic Response Surface Method) suitable for solving practical design optimization problems. The estimates based on a model proposed from optimization were in accord with the results of the experiment and road noise reduction was achieved by applying the optimally designed lower arm of the vehicle suspension system to a real vehicle.

As advances in automotive design and manufacturing technology have emerged, the demand for noise reduction and ride quality improvement has increased. In particular, under normal driving conditions, drivers are most exposed to road noise, which can be unpleasant. Furthermore, it can reduce ride quality and brand awareness. Noises within the passenger compartment are categorized as structural and air-borne noises, depending on the transfer path characteristics. Road noise is the most common structural noise and is caused by road excitation to the tires in the medium- to low-frequency range (0–500 Hz). This energy is transferred to the vehicle suspension system [1]. Thus, each component of the vehicle suspension system should work to isolate against road vibrations, this requires various studies to improve vehicle suspension systems and reduce road noise [2], [3] and [4]. Road noise transferred to the passenger compartment has a complex vibration mechanism and multiple vibration sources, making it difficult to develop precise countermeasures. In particular, when road vibration is transmitted to the vehicle suspension system and car frame, the vibration generated from each element in the transfer path, as well as the specific transfer path characteristics, is modeled in multiple input and single output systems to perform effective noise reduction strategies using contribution estimations. Moreover, it is critical for the designer to have detailed information on vibration sources to reduce noise in the passenger compartment. This information will help save a lot of time and effort compared to the traditional method of relying on experience and intuition. In this respect, this study uses sensitivity analysis to develop a new vibration isolation design in order to identify the relationship between design variables and objective functions [5] and [6]. In terms of design sensitivity, Haftka and Adelman reported results on design sensitivity in discrete systems up to the late 1980s [7], whereas in 1968, Fox and Kapoor reported changes in the eigenvalue and eigenvector of the symmetric matrix [8]. C.C. Hsieh and J. S. Arora set up a first-order differential equation for mechanical structures to present a sensitivity analysis method using direct differentiation and an adjoint variable method [9]. However, these methods require a precise understanding of the design object and complicated kinematic equations. Thus, FE analysis is needed for simpler applications. This study performed a vehicle driving test to analyze road noise characteristics and identified the frequency appropriate for noise reduction. Moreover, this study also developed a model of road noise delivered to the passenger compartment using multiple input and single output systems to quantify coherent output estimates through transfer path analysis (operation transfer path analysis: OTPA). For components that serve as the vibration source, we performed finite element analysis using Abaqus and sensitivity analysis using direct differentiation [10] and [11] to identify important design variables. To obtain the optimum values of the important design variables identified, we employed a robust and efficient sequential approximate optimization method, named PQRSM (Progressive Quadratic Response Surface Method), suitable for solving practical design optimization problems [12]. Then, we applied an optimally designed lower arm of the vehicle suspension system to a real car to assess noise reduction and compared the estimates to the actual test results.

This study identified the vibration sources of the vehicle suspension system with superior coherence and proposed an optimal design to reduce road noise that is structure-borne noise and obtained to the following conclusions. 1) Transfer path and coherence analysis was performed on the vehicle suspension system, which is the major transfer path. The study also identified vibration sources that cause road noise that reaches the passenger compartment. 2) Sensitivity analysis was performed on an FE model to identify design variables that are sensitive to object function. 3) Finally, this study proposed an improved strategy for noise reduction using design optimization. When we applied the optimally obtained design variable values to the vehicle test, we obtained results similar to those obtained by design optimization, which validate our proposed strategy.